PLATFORMS &
PROJECTS
The know how of everything we do at Trilogy Innovations.
OUR EXPERTISE
We constantly push to break ground with the latest and the most sophisticated of technologies. Hands-on involvement with leading-edge technologies ensures that we stay ahead of the curve.
PLATFORMS
TI created DevFlows, a platform for application development and Integration Platforms which enables people to build applications that integrate multiple systems easily. It provides a SaaS application, along provides a visual Flow builder interface and finally analyzes the flow.
What does it do?
DevFlows is “Visual Basic” for AWS. It is an application development and integration platform that allows you to stitch AWS and third-party services together to add new cloud-native capabilities to products. DevFlows is event based. Unlike “Visual Basic,” DevFlows requires no runtime environment other than AWS. The DevFlows platform allows you to create custom nodes/adapters for any application that you might want to make available on the platform.
What problem does it solve?
Devflows enables people who are not experts in Cloud architecture and deployment to build applications that integrate multiple systems easily. It is challenging to get integrations with third party systems right at scale. Each has its own Interfaces and quirks. It starts with picking the right API, authenticating, and handling rate limits and ends up consuming an organization’s time and energy. It takes a team of experts to pull this off without Devflows.
How does it work?
- Provides a SaaS application that you just log in to without any setup, and start building Flows. Provide a large discoverable library of useful reusable components.
- Provides a visual Flow builder interface. Simply drag, drop, configure, and deploy.
- Devflows analyzes the Flow, determines an optimal deployment architecture following best practices, and deploys your Flows.
What better way to manage the technical debt of source code, than a CAP Platform? The platform enables the management of the technical debt of source code.
What does it do?
Helps engineering organizations manage technical debt at scale
What problem does it solve?
Managing technical debt of source code.
How does it work?
The service is using a multi-strategy approach: rich analytics to gain deep insights into what technical debt needs to be managed, behavioral modeling of developers to arrest technical debt inflation and automatic fixing of technical debt issues at scale.
Where is it used?
The service is used by large engineering organizations that want to gain back control of their codebases code quality by dramatically lowering the cost of managing technical debt.
How does it simplify the problem?
By using innovative strategies, managing technical debt is now possible at an unprecedented scale for large engineering organizations.
At Trilogy, we work only on the toughest business problems. Like everything else, we see machine learning as just another tool that can be applied to solve a business problem. However, what is really fascinating is the wide variety of problems that can be solved using machine learning. We construct algorithms that perform analysis and learn based on existing data.
What does it do?
Machine Learning helps in solving the toughest business problems efficiently and quickly by using different models.
What problem does it solve?
Machine Learning helped customer support teams to derive meaningful insights, it has also enabled cement manufacturers to improve their product quality, what can be better than Saving hundreds of hours for manual annotation effort, by automatically and accurately annotating data, this is another project which made machine learning important for every organization. Some more projects that we have done using ML are :
-
Detecting frauds in video interviews using vision based activity recognition models
-
Helping telcos run automated what-if scenarios to identify the best pricing plans for their customers and to reduce customer churn at the same time.
-
Optimizing supply chain distribution by accurately predicting the future inventory demand using neural networks based forecasting models.
How does it work?
The service is using a multi-strategy approach: rich analytics to gain deep insights into what technical debt needs to be managed, behavioral modeling of developers to arrest technical debt inflation and automatic fixing of technical debt issues at scale.
How does it work?
Machines can think and it’s true. In Order to make computers think, we make programs that explain to computers how to learn by themselves to perform specific tasks. Depending on the problem domain we pick the right ML technologies such as machine learning uses models like Topic Modeling, regression models, vision and natural language models, forecasting models and use the appropriate mix of supervised or unsupervised learning techniques to solve the problem at hand.
PROJECTS
TI managed to increase the productivity of 25+ teams and created a system for ESW capital that helped in the factory mode of operation.
Why did you start the project?
Managers in ESW capital were doing a lot of garbage project management work which left them little time to focus on coaching and deep dives. There was no stock workflow in place that could be applied to all the systems (finance, org builder, engineering, import, etc.)
What have you completed?
A worker-oriented system that provides a factory mode of operation, focused on increasing the productivity of teams. The platform provides a stock workflow that is optimised for quality. It forces the process owners to define their processes in a structured manner, which enforces the assembly line mentality and helps individuals and managers gain valuable insights to keep improving the productivity of their teams. It enables the management to roll out and see the reports in days (not weeks, or months), helping them experiment faster. It also ensures that every team has a simple process/workflow with no back-loops, thereby preventing iterations.
What have you achieved?
On-boarded 25+ teams, (re)designed 150+ work units. All teams benefit significantly from the stock features of the platform, such as defined prioritization between different task types, stock workflow focusing on quality, stock coaching, stock RCA (deep dives), the extension to fetch the next task, etc. The managers agree that ALP has removed the garbage project management work allowing them to focus on coaching and deep dives. The processes have been redesigned for the better and the teams agree that it’s not just in theory, but they’re actually seeing substantial improvements after shifting to ALP. Though it is hard to quantify, we would say that most teams have improved their process by 1.5x – 2x, while the finance teams have got up to 5x efficient as they used ALP to define, simplify and optimize their processes.
Trilogy Innovations used Database, created a project named as ScaleArc, where we successfully built a DB migration tool and helped several companies to migrate their database.
Why did you start the project?
I used to be interested in Databases, as they are a very crucial part of any software system. And I thought that the “Scalearc DB migration project” will be a way for me to get introduced to the core of Database engines and stuff.
What have you completed and achieved?
I have successfully built a DB migration tool, which can migrate databases of 3 engines (MySQL, MSSQL and PGSQL) from on-premise to Cloud with just a single request. This can make the lives easier for several companies out there, by seamlessly migrating their legacy databases, which could have been rather painful with the solutions existing currently in the market.
TI helped Crossover in finding cheating instances and sensitive data and made their work uncomplicated. We have covered 155+ incidences out of 195.
Why did you start the project?
Crossover conducts 50,000-70,000 online hiring tests per week with about 10,000 candidates taking tests simultaneously. The Customer currently proctors tests manually on zoom. The current solution does not provide automated post-examination video analysis for cheating detection. This will be immensely beneficial for the customer for proctoring tests.
What have you completed?
Annotation of cheating instances and sensitive data. Tailored annotation for each type of video stream. Post Exam AI analysis on recordings based on annotation captured.
What have you achieved?
Last year, there were 365 cheating related instances that were captured here by the crossover team. With the above scope, we cover 155+ incidents out of 195 incidents that occur during exam post room sweep. [ ~ 80% incidents]
“ We create the perfect blend of innovation and state-of-the-art technologies. ”
Anuja Sivaram, Trilogy Innovations